2026-08论文中文摘要
AI控制科学家：用于自动控制设计的大语言模型驱动的智能体系统
AI Control Scientist: LLM-driven Agentic System for Automated Control Design
arXiv 2608.26780 · 2026-08-28指令质量至关重要：优化指令以实现有效的偏好学习
Instruction Quality Matters: Refining Instructions for Effective Preference Learning
arXiv 2608.26779 · 2026-08-28观测单光子的相对论轨迹
Observing relativistic trajectories of single photons
arXiv 2608.26778 · 2026-08-28面向中小企业云数据保护的混合后量子加密架构，具备自托管密钥管理功能
A Hybrid Post-Quantum Encryption Architecture with Self-Hosted Key Management for SME Cloud Data Protection
arXiv 2608.26777 · 2026-08-28超越傍轴近似的电磁光束的双对称构造：标量修正与矢量修正的显式分离
Dual-Symmetrized Construction of Electromagnetic Beams Beyond the Paraxial Approximation with an Explicit Separation of Scalar and Vectorial Corrections
arXiv 2608.26776 · 2026-08-28无环约束下合取查询的大小界
Size Bounds for CQs Under Acyclic Constraints
arXiv 2608.26775 · 2026-08-28导数层级作为高斯过程重建哈勃参数时核依赖趋势的起源
Derivative hierarchy as the origin of kernel-dependent trends in Gaussian process reconstructions of the hubble parameter
arXiv 2608.26774 · 2026-08-28平衡态非线性摩擦下的布朗运动却非高斯扩散
Brownian yet non-Gaussian diffusion through equilibrium nonlinear friction
arXiv 2608.26773 · 2026-08-28经过指定顶点的环的填充与覆盖
Packing and Covering Cycles Through Prescribed Vertices
arXiv 2608.26772 · 2026-08-28面向茶叶病害分类的、从视觉基础模型到轻量视觉状态空间模型的跨架构知识蒸馏
Cross-Architecture Knowledge Distillation from a Vision Foundation Model to a Lightweight Visual State Space Model for Tea Leaf Disease Classification
arXiv 2608.26771 · 2026-08-28ℂ^⊕ⁿ上的弱双代数结构
The weak bialgebra structures on $\mathbb{k}^{\oplus n}$
arXiv 2608.26770 · 2026-08-28沿加热壁面推进的重力流锋面
Gravity current fronts advancing along a heated wall
arXiv 2608.26769 · 2026-08-28金星大气热结构研究：VEX/Akatsuki与Venus-GRAM/VCD的对比分析
Study on the Venusian Atmospheric Thermal Structure: A Comparative Analysis between VEX/Akatsuki and Venus-GRAM/VCD
arXiv 2608.26768 · 2026-08-28全球变暖下强迫ENSO变化的可检测性：来自充放电振子（RO）的见解
Detectability of Forced ENSO Changes under Global Warming: Insights from the Recharge Oscillator
arXiv 2608.26767 · 2026-08-28MeshPriorDiT：面向动作条件布料动力学的分层建模
MeshPriorDiT: Hierarchical Modeling for Action-Conditioned Cloth Dynamics
arXiv 2608.26766 · 2026-08-28超越客户端平均：随机SCAFFOLD中的客户端无关二阶稳态偏差分量
Beyond Client Averaging: A Client-Independent Second-Order Stationary-Bias Component in Stochastic SCAFFOLD
arXiv 2608.26765 · 2026-08-28用于渗流的神经重整化群流
Neural Renormalization Group Flow for Percolation
arXiv 2608.26764 · 2026-08-28用于折扣和收益的分类器自动机
Categorizer Automata for Discounted-Sum Payoffs
arXiv 2608.26763 · 2026-08-28JWST与Roman时代的经典造父变星：新理论框架及其金属丰度依赖关系
Classical Cepheids in the JWST and Roman Era: A New Theoretical Framework and its Metallicity Dependence
arXiv 2608.26761 · 2026-08-28西格尔上半空间伯格曼空间上的紧性与本质范数
Compactness and the essential norm on Bergman spaces of the Siegel upper half-space
arXiv 2608.26760 · 2026-08-28密集仓库中在线多智能体取送货任务的固定安全港预留机制
Fixed-Haven Reservation for Online Multi-Agent Pickup and Delivery in Dense Warehouses
arXiv 2608.26759 · 2026-08-28HOLMES：用于高良率估计的上下文内失效中心定位
HOLMES: In-Context Failure-Center Localization for High-Dimensional Yield Estimation
arXiv 2608.26758 · 2026-08-28DEEPCHART：大型语言模型在忠实的数据科学图表生成方面存在多大差距？
DEEPCHART: How Far are LLMs from Faithful Data-Science Chart Generation?
arXiv 2608.26757 · 2026-08-28设计安全：带连续动作的上下文博弈机的实现代价约束
Safety by Design: Realized-Cost Constraints for Contextual Bandits with Continuous Actions
arXiv 2608.26755 · 2026-08-28STREAM：面向工业能源数据采集的目标驱动且感知不确定性的框架
STREAM: An Objective-Driven and Uncertainty-Aware Framework for Industrial Energy Data Acquisition
arXiv 2608.26754 · 2026-08-28超越执行：审计大语言模型驱动科学研究中的实验保真度
Beyond Execution: Auditing Experimental Fidelity in LLM-Driven Scientific Research
arXiv 2608.26753 · 2026-08-28基于3D视觉几何的玻璃表面检测
Glass Surface Detection Grounded in 3D Visual Geometry
arXiv 2608.26752 · 2026-08-28Kagome金属中电荷有序波矢的层间工程
Interlayer-engineering of Charge Order Wave Vector in Kagome Metals
arXiv 2608.26751 · 2026-08-28利用大语言模型发现数据湖中的关系：一个工业案例
Discovering Relationships in Data Lakes Using Large Language Models: An Industrial Case
arXiv 2608.26750 · 2026-08-28字母将真相隐藏于我们眼前：英语同音异义词具有显著不同的语音实现形式
Letters hide the truth from our eyes: English homophones have meaningfully different phonetic realizations
arXiv 2608.26749 · 2026-08-28用于物理感知生成的自增强扩散引导
Self-Augmented Diffusion Guidance for Physics-Informed Generation
arXiv 2608.26748 · 2026-08-28随机形变高斯酉系综最大特征值的中偏差
Moderate Deviations for the Largest Eigenvalue of a Randomly Deformed Gaussian Unitary Ensemble
arXiv 2608.26745 · 2026-08-28G2D：用于零样本图像分类的生成式到判别式协同推理框架
G2D: Generative-to-Discriminative Collaborative Inference for Zero-Shot Image Classification
arXiv 2608.26744 · 2026-08-28面向语言模型的图引导选择性遗忘：控制遗忘种子之外的支持路径
Graph-Guided Selective Unlearning for Language Models: Controlling Support Routes Beyond Forget Seeds
arXiv 2608.26743 · 2026-08-28Claude Code 完整用户手册
Claude Code Complete User Handbook
arXiv 2608.26742 · 2026-08-28从一般更新-跳跃动力学推导亚扩散方程
Deriving subdiffusion equations from general Renewal--Jump dynamics
arXiv 2608.26740 · 2026-08-28基于残差深度强化学习的缆绳驱动下肢康复机器人计算力矩控制（针对扰动与参数不确定性场景）
Residual Deep Reinforcement Learning-Based Computed Torque Control for a Cable-Driven Lower-Limb Rehabilitation Robot under Disturbances and Parametric Uncertainties
arXiv 2608.26739 · 2026-08-28二维非相对论交替磁体的不稳定性
Instability of two-dimensional nonrelativistic altermagnets
arXiv 2608.26738 · 2026-08-28生成式语义场景补全
Generative Semantic Scene Completion
arXiv 2608.26737 · 2026-08-28动态树颜色：具有最小不稳定性的自适应可区分层次结构
Dynamic Tree Colors: Adaptive Discriminable Hierarchies with Minimum Instability
arXiv 2608.26734 · 2026-08-28白日梦：通过黑盒任务交互窃取隐藏的智能体技能
Daydreaming: Stealing Hidden Agent Skills through Black-Box Task Interaction
arXiv 2608.26733 · 2026-08-28将消息传递重新思考为用于文本属性图学习的检索
Rethinking Message Passing as Retrieval for Text-Attributed Graph Learning
arXiv 2608.26732 · 2026-08-28J0815+4729中的碳：对最原始恒星中CH吸收的解释投下质疑的阴影
Carbon in J0815+4729: Casting A Shadow of Doubt on the Interpretation of CH Absorption in the Most Primitive Stars
arXiv 2608.26731 · 2026-08-28知晓何时不重用：自主大语言模型后训练中的条件经验迁移
Knowing When Not to Reuse: Conditional Experience Transfer in Autonomous LLM Post-Training
arXiv 2608.26730 · 2026-08-28超越单一视角：针对推荐系统中稀疏且不完整的用户生成内容的元评审
Beyond a Single Story: Meta-Reviewing Sparse and Incomplete User-generated Contents for Recommendation
arXiv 2608.26728 · 2026-08-28面向超大规模多输入多输出（XL-MIMO）的摊销神经奇异值分解：用于波束成形和多流效用的结构引导因子预测
Amortized Neural SVD for XL-MIMO: Structure-Guided Factor Prediction for Beamforming and Multi-Stream Utility
arXiv 2608.26727 · 2026-08-28近视界BMS对称性及其对黑洞熵的启示
Near-Horizon BMS Symmetry and Implications on Black Hole Entropy
arXiv 2608.26725 · 2026-08-28GeoMAD：基于可变形融合与分布对齐的几何感知多视图异常检测
GeoMAD: Geometry-Aware Multi-View Anomaly Detection via Deformable Fusion and Distributional Alignment
arXiv 2608.26724 · 2026-08-28弯曲流形中特征值调和均值的外蕴上界
Extrinsic upper bounds for the harmonic mean of eigenvalues in curved manifolds
arXiv 2608.26723 · 2026-08-28UniGeo：用于文本引导跨视角地理定位的多模态大语言模型
UniGeo: A Multi-modal Large Language Model for Text-Guided Cross-View Geo-Localization
arXiv 2608.26722 · 2026-08-28
